Dothan Review: Powerful and low-power

[Xeon]

We are talking maximum thermal output and the numbers match up with the P4's so what changes for the A64's?

Min P-state 800MHz VID_VDD 1.10v IDD Max 15.3A
Apply the formula it comes to 16.83watts for the mobile processor at 800MHz.
Min P-state 800MHz VID_VDD 1.40v IDD Max 45.6A
Apply the formula again it comes to 63.84watts for the A64 3000+.

I am still not seeing what you’re getting at though, since those numbers fall into AMD’s thermal design power limits of 19watts for the mobile processors at min state and 89watts for the desktop processors at min state.

Considering a FX at 2.4 puts out 86.1watts at max thermal output and has an 89watt thermal design power limit. With margin of error I don’t really see the processors putting out much less than that perhaps at best case 20watts less for the FX line. But that’s being overly generous IMO.

With considerations that a 3.6 P4E at core voltage of 1.4v and power draw of 115watts puts out 161watts, that type of heat output is quite realistic, so I find it almost charming that you seem to believe A64’s don’t put out 60watts all full max thermal load. But either or you don’t have to believe the numbers but with the P4E’s leaking near 200 uA. I would almost say the estimates are generous.

And those are for total energy. Thermal output should be at least 15% less.

A formula to back that up might convince me otherwise, but since I pulled 15% out of my ass to begin with I will be eagerly awaiting the results.

Xeon

<font color=red>Post created with being a dickhead in mind.</font color=red>
<font color=white>For all emotional and slanderous statements contact THG for all law suits.</font color=white>

 

[trooper11]

...um and my link says what?

 

[Xeon]

The mobile processor yes yes I know it's thermal specifications. I was still argueing that the desktop models put out their fair share of heat as well.

So I think this is all cleared up perhaps another thread?

Xeon

<font color=red>Post created with being a dickhead in mind.</font color=red>
<font color=white>For all emotional and slanderous statements contact THG for all law suits.</font color=white>

 

[P4Man]

You're comparing apples to lightbulbs. AMD's specs span the entire product range, probably including upcoming 2.6 GHz 130nm parts. Is it surprising a 2.6 GHz 130nm desktop part has a considerable higher TDP than a 90nm 2 GHz mobile part ?

To compare at least apples to oranges, compare AMD's mobile A64 to Dothan, and you'll see that the mobile A64, using your math consumes up to:
27 * 1.2V = <b>32.7W @ 1.8 GHz</b>

If you consider that number includes FSB and dual channel memory controller, and its a TDPmax unlike intels TDPtyp I think its fair to say it is not in a different league as the P-M in TDP per performance, in spite of the fact this chip wasn't designed from the ground up for low power, and the fact its built on 130nm instead of 90nm.

I stand by my point, a 90nm hammer ought to be pretty competitive with Dothan at least in maximum power consumption for a given performance level, and I expect it to lead bya considerable margin in absolute performance. Average/minimum power consumption, I expect Dothan to lead still, making it a mobile chip of choice if battery life is a major concern.

= The views stated herein are my personal views, and not necessarily the views of my wife. =

 

[P4Man]

> Heat output will always be less than the consumption of
>energy

Ahm.. if the energy you put in is not converted into heat, then into what exactly ? its not like there is energy poured into potential energy, deformation energy or light or something. I think its safe to assume >99% of the electrical energy you put in is transformed into heat. If not, into what then ?

= The views stated herein are my personal views, and not necessarily the views of my wife. =

 

[trooper11]

ok now that makes more sense lol

 

[darko21]

Ahm.. if the energy you put in is not converted into heat, then into what exactly ? its not like there is energy poured into potential energy, deformation energy or light or something. I think its safe to assume >99% of the electrical energy you put in is transformed into heat. If not, into what then ?


I think Schmide has got you on this one. Enegry is a strange thing if you have enough you can create matter. energy is released in many forms to produce the desired effect. by products can be any # of things light, sound, heat probably many more things, its never close to 100% effecient. You could argue how much light a cpu gives off but then how much noise does a tree make in a forest if nobody is around to hear it? I do not know lots on this topic probably very lttle but Schmide stating "Heat output will always be less than the consumption of energy" makes perfect sence to me. A fuel injected car is probabbly only 30% effecient a lot of the wasted energy is heat noise etc I would not be too suprised if a cpu lost a fair bit of energy trying to accomplish it's purpose.


If I glanced at a spilt box of tooth picks on the floor, could I tell you how many are in the pile. Not a chance, But then again I don't have to buy my underware at Kmart.

 

[P4Man]

> I would not be too suprised if a cpu lost a fair bit of
>energy trying to accomplish it's purpose

Energy isn't lost, its transformed into some another form. AFAIK, CPU's don't radiate light, not a chemical reaction that "stores" it, no potential (height) energy, not any half significant radiation energy (which ultimately gets transformed into heat again anyway), not deformation energy and there is matter that is created either.. I'm really at a loss what else than heat it could be converted into.

> A fuel injected car is probabbly only 30% effecient

Yes, but unlike what Schmide claimed, electrical heaters are 100% efficient, no energy is wasted through ducting or combution, unlike gas/fuel heaters that may achieve his claimed 85% efficiency (off course generating and transporting electricty is far from 100% efficient, but the heater itselve *is*).

A CPU is just a fancy electrical heater, no more no less...

= The views stated herein are my personal views, and not necessarily the views of my wife. =

 

[darko21]

Energy isn't lost, its transformed into some another form.

Yes exactly energy is energy. and its the transformation (that is really what energy is) where it gets lost producing unwanted by products like heat light sound. Honestly I do not know if cpu's give off light but i'd be very suprized if they did not, one might not be able to see the light cause its encased in silicone or it's very small but there none the less same with sound. Nothing is 100% effecient if it was we might get perpetual motion.

Re: Yes, but unlike what Schmide claimed, electrical heaters are 100% efficient,

Personaly I doubt it. Even a electrical resistance eliment gives off light and I'd bet some noise on a minute level. OK OK don't get the eliment so hot so no light or less light and less heat, But what about infred? light waves come in many forms. Its a kind of light wave a human cannot see, but it with the right equipment they can from many miles away and do not try and tell me the heat travels all that way. So that is wasted energy.

I am out of my eliment here but I personaly doubt the only wasted energy by product of a cpu is heat.

If I glanced at a spilt box of tooth picks on the floor, could I tell you how many are in the pile. Not a chance, But then again I don't have to buy my underware at Kmart.

 

[P4Man]

>Honestly I do not know if cpu's give off light but i'd be
>very suprized if they did not,

LMAO ! okay, for laughs and giggles, let's assume this is true, and they do radiate light. Two things:

1) Since I remember cpu's that didn't require heatsinks, i can positively attest a CPU does not generate *anywhere near* the light of a 5 milliwatt LED, and even if it did, we are talking about 0,00025% which is totally insignificant
2) Light ultimately gets converted into heat, since it is not persistant. Since no light escapes the cpu casing (or even PC casing if you prefer), it all gets turned into heat.

> But what about infrared?

Same story, and just as unlikely.

I do believe high frequency cpu's could generate some radiation, or electromagnetic waves (like X rays), but there the same story applies again.. these waves will bounce around and ultimately hit other elements, thereby converting their energy into heat. Granted, a fraction of that may well occur outside the cpu, or even outside the PC itselve, but you are talking about such insignificant ammounts that it really doesnt belong in a discussin about power consumption. For all practical purposes, a cpu converts 100% of the power it draws into heat. Its not like if the actual number where 99.99% that it would matter one damn.

= The views stated herein are my personal views, and not necessarily the views of my wife. =

 

[P4Man]

> but it with the right equipment they can from many miles
>away and do not try and tell me the heat travels all that
>way. So that is wasted energy.

oops, I missed this part. The answer is simple though, you can "see" heat with for instance infrared equipment, because every material emits infrared radiation (the hotter, the more). So your energy has to be converted into heat first, before it can radiate infrared radiation to become visible. As for the ammount of energy wasted, keep in mind even icecubes emit infrared radiation, everything does. It doesnt matter in this equitaion, you can not say a certain ammount of energy you put in is converted into infrared radiation and therefore, not into heat. Its the other way around, its the heat itselve that generates the radiation.

>I am out of my eliment here

Yeah me too, I took my last physics course over 10 years ago

= The views stated herein are my personal views, and not necessarily the views of my wife. =

 

[Mephistopheles]

Hm... from what I can tell, you're right... the only possible energy emissions are electromagnetic in nature, and I doubt there's a lot to that anyway!

...which is why I kind of think Schmide's comment that a space heater could only have 85% efficiency, at most, is probably not quite correct... I'm not completely sure, but I do think so.

<i><font color=red>You never change the existing reality by fighting it. Instead, create a new model that makes the old one obsolete</font color=red> - Buckminster Fuller </i>

 

[P4Man]

>..which is why I kind of think Schmide's comment that a
>space heater could only have 85% efficiency, at most, is
>probably not quite correct...

I think he is confused with combustion/gaz/oil whatever heaters where some energy (heat!) gets lost through the exhaust, and the combustion process itselve may not reach 100% efficiency (some material is left unburnt, so you waste energy). Electrical heater are ~100% efficient, they have to be..

= The views stated herein are my personal views, and not necessarily the views of my wife. =

 

[Mephistopheles]

Exactly what I think, too.

...even gas or oil heaters can be more efficient than that... I once saw a gas heater with... 99.1% efficiency. Pretty impressive. Too bad you don't always want to heat up things, 'cause that's the easiest thing to do... Car engines, for instance, don't get much beyond 20% efficiency, at most...

<i><font color=red>You never change the existing reality by fighting it. Instead, create a new model that makes the old one obsolete</font color=red> - Buckminster Fuller </i>

 

[zeezee]

Heat it is mon ami, nothing else…. Energy, whichever form it is originally in, eventually turns into heat. You roll a boulder downhill, its potential energy turns into heat by friction. If you are not as lucky as you think, it smashes your house on the way and it stops. All of its energy turns into heat with the impact.

Any material whose temperature is above absolute zero radiates photons. Some people call it light. Unless the material is hot enough, you don’t see the light but it’s there. Others call it Infra-Red.

All CPU’s when they run have a temperature above absolute zero, therefore all CPU’s emit light (Prescott’s a little more though). Light ain’t an alternative to heat, but a result of it.

Anything that is hot emits a photon, which is eventually absorbed by a cooler thing, which increases the entropy of the universe, taking us slowly to a thermal death…

 

[P4Man]

> If you are not as lucky as you think, it smashes your
>house on the way and it stops. All of its energy turns into
>heat with the impact.

Are you sure ? I thought there was something called "deformation energy", but then my physics may be more rusty than yours...

= The views stated herein are my personal views, and not necessarily the views of my wife. =

 

[Mephistopheles]

Hm... I think that was a joke... Or at least I hope so...

In any case, eventually, there are other forms of energy. For instance, electric or magnetic fields store energy; in addition to that, several other methods of storing energy (internal energy) exist. Then there's work, which ultimately goes to "deformation energy", another kind of internal energy. So while most of it does get turned into heat, P4Man, you have a point there. Other things exist.

Then there's also the possibility that the energy, in whatever form, will also only be converted to heat far, far away. Like in a nuclear bomb: the infrared emissions are enough to incinerate and heat up anything in vicinity (i.e. several km), but while they're travelling, they're not heat...

<i><font color=red>You never change the existing reality by fighting it. Instead, create a new model that makes the old one obsolete</font color=red> - Buckminster Fuller </i>

 

[zeezee]

Pretty much… Heat is the most basic form of energy and any form of energy eventually turns into heat.

My boulder on its way can use a part of its energy to change the formation of other materials by hitting them. However, this doesn’t mean that the energy is used up or lost per se. After the impact either the boulder or the deformed material is hotter than it was before …and the hotter cools down, the cooler heats up, until their heat content comes to the same level.

If you get down to it, we can argue about the timing of a form of energy’s transforming to heat, like when you use your muscular energy to move the boulder to the top of a hill, where you have to wait until it rolls back to retransform its potential energy to back heat but these things don’t happen in the world of CPU’s. They transform all the power they suck off of the power supply to heat…

After all, this is what you initially said, No?

 

[P4Man]

>After all, this is what you initially said, No?

Yes, on the cpu thing I'm not arguing, we agree. I was wondering about the boulder example though. Like I said, I seem to recall that when materials change their structure (because of an impact for instance), this "absorbs" energy, known as deformation energy. IOW, your boulder running downhill and not hitting anything will see 100% of its energy transformed into heat, but if it hits a styrofoam block that deforms, less energy would be converted into heat, the delta is required to deform the styrofoam. Well.. that is something I seem to recall, but I could be off... either way, this obviously doesn't apply to cpu's

= The views stated herein are my personal views, and not necessarily the views of my wife. =

 

[zeezee]

> Hm... I think that was a joke... Or at least I hope so...

Man, you are breaking my heart. It’s not a joke… I am damn serious.


> In any case, eventually, there are other forms of energy.

Sure there are. But they all are transformed to heat eventually.


> For instance, electric or magnetic fields store energy; in addition to that, several other methods
> of storing energy (internal energy) exist.

You missed Enterprise’s anti-matter warp engines For the record, electric energy, electro-magnetic energy, Infra-Red radiation, Ultra-Violet radiation, etc. are all the same type of “energy” if you will. They all are based on photon radiation which is also known as electro-magnetic radiation…. But they all turn to heat eventually.


> Then there's work, which ultimately goes to "deformation energy", another kind of internal energy.

Can you elaborate on this concept of internal energy? Do you mean to say potential energy?


> So while most of it does get turned into heat, P4Man, you have a point there.

Not most of them. ALL of them


> Other things exist.

Really? What are these “Other Things”?

 

[darko21]

"I do believe high frequency cpu's could generate some radiation, or electromagnetic waves (like X rays), but there the same story applies again.. these waves will bounce around and ultimately hit other elements, thereby converting their energy into heat. Granted, a fraction of that may well occur outside the cpu, or even outside the PC itselve, but you are talking about such insignificant ammounts that it really doesnt belong in a discussin about power consumption. For all practical purposes, a cpu converts 100% of the power it draws into heat. Its not like if the actual number where 99.99% that it would matter one damn."

AFAIK energy cannot be created or destroyed it just changes form. There are many types of energy light sound mechanical chemical and of course heat. Heat seems to be a common by product when energy is being used or transformed. How sure are you on waves (encased) bouncing around hitting other atoms and converting to only heat energy. lets say you take 3 rooms totaly insulate them no light sound or heat can escape. Now put a 300 watt light in one room, 300 watt heating eliment in the other and a 300 watt speaker in the last. Give it 24 hours then check the temps in each room. I'd be inclined to think the temps in the room using a heater would be much higher than the other 2 with the sound room being the coolest.

Regardless I am sure most of the wasted energy in a cpu comes out as heat.I'm just suprized to think its anywhere near 99.9% heat as a by product.

If I glanced at a spilt box of tooth picks on the floor, could I tell you how many are in the pile. Not a chance, But then again I don't have to buy my underware at Kmart.

 

[Mephistopheles]

Erm, no, not 100% of all energy in the universe will eventually be converted into heat. Take, for instance, the energy that binds nuclei together. It can be released by a nuclear reaction, freeing energy in the form of heat, but before that, it's not heat.

For the record, electric energy, electro-magnetic energy, Infra-Red radiation, Ultra-Violet radiation, etc. are all the same type of “energy” if you will. They all are based on photon radiation which is also known as electro-magnetic radiation….

According to quantum field theory, you are right. However, I was merely stating that photons are energy transmitters and as such aren't heat. And not all energy gets transformed into heat at all. I mean, if you carried your boulder up a hill, energy gets transformed into potential energy, not heat. As long as you don't roll the boulder down, it remains stored as internal energy.

Can you elaborate on this concept of internal energy? Do you mean to say potential energy?

Internal energy refers to any kind of energy storing process that happens inside a system. Classically, heat would be one way to do that, but there are other kinds of internal energy as well, like potential energies, if you will. Ways to store energy exist aside from heat; an endothermal chemical reaction, for instance, actually decreases temperature. And ultimately, mass is also energy, but certainly all mass will not turn into heat.

What is obviously true is that when dealing with computer chips, all of the energy consumed will eventually turn into heat in the end, but this is not at all true for all processes. What if you emit a photon into space? It won't get converted into heat, and it's definitely energy.

<i><font color=red>You never change the existing reality by fighting it. Instead, create a new model that makes the old one obsolete</font color=red> - Buckminster Fuller </i>

 

[P4Man]

>AFAIK energy cannot be created or destroyed it just changes
>form.

Exactly.. now think of under what form a CPU would generate other energy than heat..

> How sure are you on waves (encased) bouncing around
> hitting other atoms and converting to only heat energy

Fairly certain, unless those waves would somehow be able to alter to molecular structure; which could "absorb" energy or even cause more energy to be released (think nuclear power). Neither is likely to happen inside a cpu though If it helps you grasp this, consider radar waves, and try this: go to a military airport, stand in front of the radar cone of a plane, ask the pilot to turn on its radar, and you'll feel exactly what I mean (you'll get literally fried by the microwaves).

> lets say you take 3 rooms totaly insulate them no light
>sound or heat can escape. Now put a 300 watt light in one
>room, 300 watt heating eliment in the other and a 300 watt
>speaker in the last. Give it 24 hours then check the temps
>in each room. I'd be inclined to think the temps in the
>room using a heater would be much higher than the other 2
>with the sound room being the coolest.

Assuming the speaker actually draws 300 watt of electrical power (wattage of speakers is kinda fuzy, depends on impedence and frequency,..), they will be exactly as hot/cool.

>Regardless I am sure most of the wasted energy in a cpu
>comes out as heat.I'm just suprized to think its anywhere
>near 99.9% heat as a by product.

Not sure if its "by product", its the only thing a cpu produces with the energy its fed. If anything is amazing, then its the enormous amount of current a cpu requires (>100 ampere for Prescott!) just to operate..

= The views stated herein are my personal views, and not necessarily the views of my wife. =

 

[zeezee]

> Erm, no, not 100% of all energy in the universe will eventually be converted
> into heat. Take, for instance, the energy that binds nuclei together.
> It can be released by a nuclear reaction, freeing energy in the form
> of heat, but before that, it's not heat.

We are going off-topic but whatever. There is no difference between my boulder which I brought up the hill against gravity (which increased its potential energy by moving it apart from the center of a gravitational radiation) and your nuclei where protons and neutrons stay apart and don’t collapse against the gullion or z-particle radiations. It’s all potential energy. Drop them in a black hole, you get your heat back when they are smashed (and please don’t start like Hawking, Wormholes, Black Holes don’t radiate,, etc).

> According to quantum field theory, you are right.
> However, I was merely stating that photons are energy
> transmitters and as such aren't heat. And not all energy
> gets transformed into heat at all.

Heat is vibration of molecules. Photons are sub-atomic particles and therefore are not and can not be “heat” anyway. They just transfer energy between atoms and molecules and make them vibrate faster/slower. But please come back to our mortal and corporeal continuum. We are talking about CPU’s. No?

> I mean, if you carried
> your boulder up a hill, energy gets transformed into
> potential energy, not heat. As long as you don't roll the
> boulder down, it remains stored as internal energy.

Jeez. Read my previous post. There is always the timing issue but not with the CPU’s. If you fuse the atoms of a Prescott, you get much more than 70 C. And this isn’t gonna happen until the universe collapses back for the next big-bang.

Anyway, let’s cool down. I have the feeling that we are boring the audience with things that they haven’t come here to hear about.

 

[darko21]

Re: "If it helps you grasp this, consider radar waves, and try this: go to a military airport, stand in front of the radar cone of a plane, ask the pilot to turn on its radar, and you'll feel exactly what I mean (you'll get literally fried by the microwaves)."

I'm not suprized but I don't see how that is relevant. I'll take your word I might want kids one day.


Re: "Assuming the speaker actually draws 300 watt of electrical power (wattage of speakers is kinda fuzy, depends on impedence and frequency,..), they will be exactly as hot/cool."

So this sound like you are saying heat output is a constant when energy is transformed from any one type to any other type of energy or at least electrical energy. Common sense tells me a room will heat up more using a 300 watt heater over a speaker drawing a constant 300 watts. But I could be wrong. I'd just like somone to explain how 300 watts going in always produces the same heat output take a 300watt air conditioner sure it creates some heat in the process but not the heat a 300watt heater would.


If I glanced at a spilt box of tooth picks on the floor, could I tell you how many are in the pile. Not a chance, But then again I don't have to buy my underware at Kmart.